Who and in which schools’ children can be promoted to the next grade in Malawi? From longitudinal data

• In Event: Education access and inequality around the globe

Proposal

Sustainable Development Goals 4 targets all children complete quality primary education by ensuring basic learning outcomes. However, UNESCO (2020) reported that approximately 19% of the children are out of primary school, and only 65% completed primary school in Sub-Sharan Africa in 2018. Reducing grade repetition and dropout rates are urgent issues. This study investigates who and children in which school can be promoted to the next grade from longitudinal data for rural Malawi. To pursue that aim, the study addresses three research questions: (1) What is the proportion of students promoted to the next grade? (2) What kinds of individual and family factors influence students' grade promotion? (3) What kinds of teacher and school factors influence students’ grade promotion?
The study used the researcher's framework based on Scheerens' school effectiveness research (Scheenens, 1990). It is a context-input-process-outcome model of schooling in the functioning of the educational system. The feature of this study is to collect data through survival analysis. Survival analysis is a method that identifies the causes of an event at the time. The same students are followed to capture students promoted to the next year, repeat grades, transfer to another school, and drop out of school. Then, this study explores the causes of grade promotion. By doing so, it can identify the characteristic of promoted students and characteristics of school children commute.
This study selected the Nkhata Bay District in Malawi as a research area. By using a stratified random sampling method, 30 public primary schools were selected based on the pass rates of national examination during the past three years. The field research was undertaken from 2013 to 2015. At the baseline of the study in 2013, the study collected data from 1745 grade 5 students from the selected schools. In 2014, the activity was undertaken upon revisiting the same schools in 2013: Confirming whether students had been promoted to the next grade, repeated the same grade, transferred to another school, or dropped out of school. In 2015, the same activity in 2014 was undertaken. Students at the baseline in 2013 followed for three years by checking whether they are promoted to the next grade.
Two types of research instruments were used: achievement test for English and mathematics as well as questionnaires. The researcher designed the tests' framework to measure vocabulary, grammar, conversation, and reading comprehension in English, and counting and writing numbers, addition, subtraction, multiplication, division, shapes, patterns, fractions, and graphs in mathematics. The researcher developed tests in both 2013 and 2014 with reference to curricula, textbooks, zonal and national examinations, and SACMEQ assessments. The English and mathematics tests each consisted of 35 question items. To enable to compare tests in 2013 and 2014, the researcher used equating in Item Response Theory by including the same question items to the two test forms: 16 question items for English and mathematics tests in grades5 and 6. Three types of questionnaires (for students, head teachers, and teachers) were developed by referring to literature on grade repetition, transfer, and dropout in developing countries. In addition, the questionnaires from a SACMEQ study were used as a follow-up because so few studies on grade promotion, grade repetition, dropout, and transfer have been conducted in low income countries, which are measured variables is limited. The questionnaire for students was translated into the local language, Chichewa, to prevent language difficulties from affecting the results. Back translation was carried out to check for translation mistakes. The English and mathematics tests were distributed to students, and questionnaires were distributed to students, headteachers, and teachers in the selected school in both 2013 and 2014. In the analysis, multilevel Cox-regression was selected because data is nested as it is students in school. The five time-constant variables at the student level were used: Gender, preprimary experience, number of grade repetition, repetition at grade 5, and age of first entry. The nine time-vary variables at the student level were used: Achievement, number of absent days, health condition, educational aspiration, self-esteem, feeling about school, parents' educational level, family economic status, and number of siblings. The ten time-constant and time-vary variables at the school level were used: Class teacher number of in-service training courses have attended, length of teaching, number of absent days, class size, school location, school facilities, length of being headteacher, training school management, handing teacher behavior, co-operation from the community. The three models were created: Model 1 for the time-constant variables at the student level, Model 2 for the time-constant and time-vary variables at the student level, Model 3 for the time-constant and time-vary variables at the student and school levels.
Grade promotion rates were significantly low. From 2013 to 2014, only 707 of 1542 students (45.8%) were promoted to the next grade. From 2014 to 2015, 293 of 707 students (41.4%) were promoted to the next grade. The results of multilevel cox regression showed that in Model 1, the age of first entry, number of grade repletion, and preprimary experience were significant factors of grade promotion. In Model 2 for adding time-vary variables at the student level, only the age of first entry remained as a significant factor, and other variables were significant: Achievement, health condition, educational aspiration, self-esteem, and feeling about school. However, in Model 3 for adding variables at the school level, only achievement, health condition, and educational aspiration remained as significant factors. In contrast, many school-level variables showed significantly: Class teacher number of in-service training courses have attended, length of teaching, number of absent days, class size, school facilities, length of being headteacher, training school management, handing teacher behavior, co-operation from the community. The results indicated that student factors and teacher and school factors influence student grade promotion. The findings contribute to ongoing discussions about the educational policy regarding improving completion rates. The methodology used in this study can also apply to studies on education transition in other developing countries.

Author

• Kyoko Taniguchi, Hiroshima University